Complex behavior relies on coordinated, multi-region neural spiking dynamics that make up the “internal conversation” of cognition. My past and current research has been dedicated to developing electrophysiological, computational and conceptual tools to characterize this conversation at a brain-wide scale and with moment-to-moment resolution.  To do this, I combine chronic recordings of thousands of neurons across the rat brain, pharmacological and genetic tools for circuit manipulation, and computational tools for inferring low-dimensional neural dynamics.

Currently: I am an Associate Research Scholar in Dr. Carlos Brody’s lab at the Princeton Neuroscience Institute. My current research focus is on an emerging mystery in the neuroscience of decision making: a subject’s choice in perceptual decision making tasks can be decoded with increasing accuracy over time from many regions throughout the brain. But how is this brain-wide activity coordinated, spatially and temporally, to generate a behavioral choice? To address this, we developed tools to simultaneously record from thousands of neurons across up to twenty brain regions in rats performing a decision making task requiring gradual accumulation of auditory evidence. This allowed decoding, on single trials and for each of the simultaneously recorded regions, the time-varying state of the decision as it is being formed.  The decision state varied across time in a way that was remarkably coherent across regions, indicating a single integrated decision computation that is shared across the brain. Simultaneous decoding of decision signals from many brain regions also allowed comparing their lead-lag relationships on single trials, revealing that they are initially computed in localized frontostriatal subnetworks and then globally propagated.

Previously: I did my PhD in Bruce Cumming’s lab at the Laboratory for Sensorimotor Research at NIH in Bethesda, MD, studying visual decision making in monkeys. At the time, the coding fidelity of the visual system was widely thought to be limited by correlated variability in neural responses (i.e. “noise correlations”). Using large-scale population recordings in primary visual cortex while monkey subjects performed a set of orientation discrimination tasks, I showed instead that correlated variability changes systematically with task, primarily reflecting downstream decision-related computations conveyed via feedback. This discovery dramatically changed the field’s interpretation of variability of sensory responses, demonstrating that, even at the earliest stages of cortical sensory processing, neural activity during decision making can only be understood in the context of multi-region recurrent interactions.